Part 5: Mixing
It can be tempting to use different reverbs for different instrument types, but what I found is that this can make things sound over-produced and less natural. I recommend using a single reverb for everything. Create an aux track with a reverb insert, and then use sends to send the signal from your track to the reverb. Then just copy that send to the rest of the tracks so they’re all the same. If someone is particularly close-mic’d you may want to raise the level of the send to get more reverb on them. In some cases a recording may have some natural reverb from the room it was recorded in and lowering the send may help.
At first I just panned everyone the same way (i.e. all the sopranos somewhat off to the left, all the alto somewhat off to the right, all the first violins had the same setting, all the 2nds had the same setting and so on). That didn’t sound bad, but it seemed like I lacked a bit of depth and dimension that we get on a scoring stage. Then I realized that in real life the musicians aren’t all in the exact same spot, they’re next to each other.
As you can see in the photo above, the musicians are spread out in their sections. So I tried mimicking this with my panning. For example instead of panning all the sopranos in the video I shared on page 1 to -60 (about 2/3rds left of center) I panned each one slightly differently ranging from -80 to -45. The difference was subtle, but to my ears it seemed to add some depth and dimension to the section. I did the same for each section.
As I discussed in Part 1, mono recordings should be delivered as mono tracks, but often people don’t know how to bounce them that way and deliver stereo tracks instead. Whenever I had those I treated them as mono tracks as you can see in the above screenshot where both panners (R & L) were panned the same.
For the most part I found there’s not a lot of need to EQ tracks I received. However on occasion a track would just stick out because certain frequencies popped out. In other cases there was some noise or hum due to a less than stellar recording – this was particularly true for some of the amateurs and the few who recorded themselves directly on their phones.
If someone was a little bright that could be toned down with EQ so they don’t stick out. If someone sounded a bit dull, they could be brightened up. If there was hum I would use a low cut or shelf to eliminate or reduce it. There’s no shortage of EQ plug-ins out there, and the built-in EQs that DAWs come with are all perfectly fine. But I prefer fabfilter Pro Q3. I love how clear and transparent it sounds. I think their visual interface is outstanding. Furthermore it includes great advanced features such as dynamic EQing and EQ matching all at a very reasonable price.
Blending multiple individual instruments so they sound like a section requires finding volume levels that work well together so no individual stand out. What worked for me was to start with a single track at the perceived level I felt was right. Then I would add the next instrument turned all the way down and slowly raise the volume until I heard it affect the first track. In my experience any more than that would cumulatively raise the volume and individual tracks kept sticking out. But by just sneaking it up until it change the sound of the first single instrument (or singer) seemed to do the trick. I repeated this process for each track. I would solo the last track I did, then add the next one. Then when I had all my tracks I played them all together and things seemed to more or less fall in place.
On occasion someone would stick out because they may have played a hairpin a bit more than the others, or played out in a particular moment more than the others. I’d hunt them down, and then simply write some automation to help blend the levels so nobody sticks out. I would ride the levels as needed on individual tracks – often it was only necessary in specific moments rather than for the entire piece.
Once everything is in place, it’s time to adjust the overall levels. I was working in ProTools so I created VCA tracks for each section so I could control the entire section with a single fader. If you prefer you can create an aux track and buss all the individual instruments within a section to that aux, it’ll achieve essentially the same end result. The final step of the mix was to adjust the levels using my VCAs to reach my final mix.